The present invention generally relates to the field of distributed computing, and more particularly to an algorithm for a client-initiated leader election in a distributed client-server system.
A distributed system may generally include numerous autonomous computational units, each of which may have its own local memory. Such autonomous computational units may communicate with each other by message passing. Each, computational unit may be typically referred to as “a computer node” or simply “a node”. In the distributed system, there is usually a shared goal which may include, for example, solving a large computational problem. In some instances, each node of the distributed system may have its own user with specific needs. In such cases, the goal of the distributed system may be to coordinate the use of shared resources or provide communication services to the users. The structure of the distributed system (network topology, network latency, number of computers) may not be recognized in advance; however the distributed system may consist of different types of computers and network links, which may vary during the execution of a distributed program.